JetClass-II dataset generation

This repo provides the generation scripts for the JetClass-II dataset.

Setup environment

Please ensure the following software is configured:

• MadGraph5_aMC@NLO and Pythia8: Download MadGraph, and install Pythia within it. Download the 2HDM model within MadGraph.
• LHAPDF: Install LHAPDF or configure an existing installation so that it can be interfaced with MadGraph.
• Delphes: Install Delphes or configure an existing installation. Link the pileup file /eos/cms/store/group/upgrade/delphes/PhaseII/MinBias_100k.pileup within this directory.

The script run.sh produces events via MadGraph + Pythia + Delphes in one go. Please configure the environment variables in the script.

Run MadGraph + Pythia + Delphes in one go

# Start the production by executing:
# ./run.sh [process_name] [num_tot_events] [num_events_per_gen_step] [job_num]

# For example:
./run.sh jetclass2/train_higgs2p 100 50 0

Expand to see full setup for JetClass-II production.

Process	Command
Res2P (neutral $X$)	./run.sh jetclass2/train_higgs2p 100000 100 [job_num], [job_num] ranges from 0–191
Res2P (charged $X$)	./run.sh jetclass2/train_higgspm2p 100000 100 [job_num], [job_num] ranges from 0–143
Res34P	./run.sh jetclass2/train_higgs4p 100000 100 [job_num], [job_num] ranges from 0–1439
QCD:	./run.sh jetclass2/train_qcd 100000 100 [job_num], [job_num] ranges from 0–239

MadGraph + Pythia step

The MadGraph + Pythia job is configured by [process_name], which takes a subdirectory path in gen_configs. The following files are present to define the job:

├── mg5_step1.dat         # Definition of the MadGraph process
├── mg5_step2_templ.dat   # Configuration for launching a MadGraph process. $1, $2, etc., serve as placeholders for parameters from a given list
├── mg5_params.dat        # Parameter list to replace $1, $2, etc.
└── py8.dat               # Pythia8 data file

Delphes step

The Delphes step is configured by a Delphes card. Three Delphes cards are provided in delphes_cards, which are derived from the CMS detector conditions in Run 2/3, adding track smearing as done in the JetClass-I production [JetClass Delphes card], and emulating pileup with $<\mu>=$50, mitigating the PU with the PUPPI algorithm.

The three JetClass-II Delphes cards have gradually increasing content and computational costs:

Delphes card	Description
delphes_card_CMS_JetClassII_onlyFatJet.tcl	This data card only produces large-R jets JetPUPPIAK8, JetPUPPIAK15 and the corresponding GEN-jets GenJetAK8, GenJetAK15.
delphes_card_CMS_JetClassII_lite.tcl	This data card also perserves small-R jets JetPUPPI (with R=0.4) and other event-level objects, including Electron, Muon, Photon, PuppiMissingET, etc.
delphes_card_CMS_JetClassII.tcl	This data card includes consistent information as in the nominal CMS card, also including the CHS jets with different radii: Jet, JetAK8, and JetAK15.

Produce ntuples from Delphes

Enter delphes_analyzers, compile, and run the makeNtuples.C macro to convert a Delphes file into ntuples. The ntuple is jet-based, with its branch definition available here.

The following example works on EL9 machines.

# Setup environment

source /cvmfs/sft.cern.ch/lcg/views/LCG_104/x86_64-el9-gcc13-opt/setup.sh
export ROOT_INCLUDE_PATH=$ROOT_INCLUDE_PATH:/cvmfs/sft.cern.ch/lcg/releases/delphes/3.5.1pre09-9fe9c/x86_64-el9-gcc13-opt/include

# Compile and run macro
root -b -q 'makeNtuples.C++("events_delphes_higgs2p_example.root", "out.root", "JetPUPPIAK8", "GenJetAK8", true)'

## defination: 
##  void makeNtuples(TString inputFile, TString outputFile, TString jetBranch = "JetPUPPIAK8", TString genjetBranch = "GenJetAK8", bool assignQCDLabel = false, bool debug = false)